Dynamic Operation Model for Power Demand Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing power demand prediction systems, such as PTL 1, fail to accurately predict power demand, resulting in significant errors, which are not adequately addressed by current methods.

Innovation Solution

A prediction system and method that utilizes a control device with a predetermined operation model to calculate prediction values by adapting to temporal attributes of data, incorporating timeliness and reliability indices to refine demand forecasts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a predetermined operation model is used to calculate prediction values, then the calculation process is simple and efficient, but the prediction accuracy is insufficient due to inability to adapt to temporal attributes of data

Engineering Contradiction:
Improveprediction calculation efficiencyVSAvoidprediction accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent applies the dynamics principle by making the operation model adaptable to temporal attributes of data. The control device changes the operation model based on temporal attributes such as timeliness and reliability indices, transforming a static model into a dynamic one that can adjust to different data characteristics and time-related factors, thereby improving prediction accuracy while maintaining computational efficiency

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent implements parameter changes by modifying the operation model parameters according to temporal attributes of the input data. The control device calculates timeliness and reliability indices and uses these to adjust the operation model parameters, allowing the same base model to adapt to different temporal conditions and improve prediction accuracy without requiring multiple complex models

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If data with strong temporal correlations is emphasized through timeliness and reliability indices, then prediction accuracy improves, but the calculation complexity increases

Engineering Contradiction:
Improveprediction accuracyVSAvoidcalculation system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing timeliness and reliability indices for the input data before performing the main prediction calculation. The control device prepares these temporal attribute assessments in advance, which then guide the operation model adjustment, reducing the complexity of the main prediction process while still achieving high accuracy

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses timeliness and reliability indices as intermediary elements that bridge the raw input data and the operation model. These indices serve as intermediate calculations that translate temporal characteristics into model adjustment parameters, simplifying the overall system architecture by creating a clear intermediate layer between data input and prediction output

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS11107094B2Prediction system and prediction method
Publication Date: 2021.08.31 HITACHI LTD
  • US11107094B2 patent drawing
  • US11107094B2 patent drawing
  • US11107094B2 patent drawing

AI summary

Provided is a prediction system for calculating a prediction value related to a prediction target to which prediction in an arbitrary period is adapted. The prediction system includes a storage device which records a plurality of data used to calculate the prediction value and a control device which includes a predetermined operation model and applies the plurality of data to the operation model to calculate the prediction value. The control device changes the operation model, on the basis of information of respective temporal attributes of the plurality of data.